Comparative and Multimodal Analysis of the EGFR, HER2, C-Myc, and MET Copy Number Alteration Using in Situ Hybridization in Korean Colorectal Cancer Patients with Integration of Array-Based Copy Number Data from the Cancer Genome Atlas

Background: The purpose of this study was to explore gene copy number (GCN) variation of EGFR, HER2, c-MYC, and MET in patients with primary colorectal cancer (CRC). Methods: Dual-colour silver-enhanced in situ hybridization was performed in tissue samples of 334 primary CRC patients. The amplification status (GCN ratio ≥2) and GCN gain (average GCN ≥4) data for the EGFR, HER2, c-MYC and MET genes were obtained. GCN variation was also assessed by the criterion of the 2013 ASCO/CAP guidelines for HER2 testing. The publicly available genetic data of 257 CRC cases from The Cancer Genome Atlas (TCGA) were used for comparative analysis. Results: Amplification of EGFR, HER2, c-MYC and MET was detected in 8 (2.4%), 20 (6.0%), 29 (8.7%), and 14 (4.2%) patients, respectively. Of 66 patients with at least one amplified gene (HER2-MET co-amplification: two patients; HER2-c-MYC co-amplification: two patients; EGFR-c-MYC co-amplification: one patient). There were 109 patients with GCN gains of one or more genes (EGFR: 11/334, HER2: 29/334, c-MYC; 60/334, MET: 48/334) and 32.1% (35/109) had multiple GCN gains. When each GCN was assessed by the criterion of the ASCO/CAP 2013 guideline for HER2 testing, 116 people showed positive or equivocal results for one or more genes. The cumulative amplification status had no association with patients’ outcome. However, the cumulative results of the GCN gain and GCN status determined according to the ASCO/CAP guideline had a significant prognostic correlation in the univariate (P values of 0.006 and 0.022, respectively) and multivariate analysis (P values of 0.010 and 0.017, respectively). In analysis of TCGA data, high-level amplification of EGFR, HER2, c-MYC, and MET was observed in 1 (0.4%), 8 (3.1%), 11 (4.3%) and 1 (0.4%) cases, respectively. Copy number alteration status had no prognostic association in TCGA cohort. Conclusions: In this study, we evaluated GCN variation of four genes in a large sample of Korean CRC patients. The amplification status was not related to patient outcome. However, the GCN gain and GCN status according to the ASCO/CAP 2013 guideline were independent prognostic factors. Legal entity responsible for the study: Individual Funding: Korea Health Industry Development Institute (KHIDI)(grant number: HI14C1813) Disclosure: All authors have declared no conflicts of interest.